The publication of these two volumes completes Neal and Cosh's corpus of all the decorated mosaics known from Roman Britain,Footnote ¹ the first complete catalogue of mosaics from any of the Roman provinces. It is a tremendous achievement, the culmination of almost ninety years' combined work on the part of the authors, going back to 1959 when David Neal painted his first mosaic at Verulamium. The statistics are mind-boggling: the four volumes of the corpus list nearly 2,000 mosaics (double the number known before the project started), from about 500 sites, with nearly 3,000 illustrations. Just to compile this much material would be a heroic task; the fact that the authors have also painted hundreds of the mosaics tessera by tessera is little short of astonishing.

Volume III covers South-East England; this is the largest volume in the series, as it incorporates the major towns of London, Verulamium, Colchester and Silchester, and the wealthy villas of Sussex and Hampshire. Volume IV covers Wales and the western counties of England, including two important centres of mosaic production at Cirencester and Gloucester; it also contains an appendix with additions to the first three volumes, listing new finds, corrections and bibliographic updates. Within each volume, the mosaics are arranged by county and then by site; brief introductions to counties and major sites outline topography, settlement patterns and history, with maps and plans showing the findspots of mosaics. Each catalogue entry systematically lists factual details — place and date of discovery, dimensions, colours, tessera sizes, approximate date, and current location (sadly, in most cases, ‘Lost’) — followed by a meticulous description of the mosaic and its setting, an account of its discovery, and a discussion of dating evidence, iconography and parallels, both within Britain and abroad. The format and level of detail are similar or even superior to other major national corpora of mosaics, such as those of France and Tunisia, which have been appearing volume by volume since the 1950s, and are also products of the drive on the part of AIEMA (L'Association Internationale pour l'Étude de la Mosaïque Antique) and its national branches to create a record of all the known mosaics from antiquity.

But what really sets the British corpus apart from its counterparts in other countries are the spectacular illustrations. Just over half the mosaics are illustrated, many with the authors' distinctive paintings, but also with photographs (especially of figured elements), line drawings, and reproductions of old engravings and paintings; the quality of the images is extremely high, and as far as possible they are in colour. Building plans are included to show the location of each mosaic; at some key sites, such as Fishbourne, Bignor, Brading and Chedworth, the paintings of the mosaics are inserted into the plan, giving an immediate visual impression of the relationship between the mosaic designs and their architectural setting. The plans, all drawn specially for the book, are admirably clear, though at the cost of simplification, as only one phase of each building is represented. Each volume also includes a pictorial glossary of motifs that appear on mosaics in the region, keyed to the multilingual dictionary of motifs in Le décor géométrique de la mosaïque romaine.Footnote ² The production quality is superlative: the illustrations are set throughout the text, the layout and typography are very handsome, and the binding should stand up to many years of use. For once, the high price is entirely justified.

The catalogue is the main object of the work, but each volume has a short introductory section which outlines the history of the region, describes the types of buildings and rooms that mosaics are found in, and gives an overview of the mosaics in the region, surveying their distribution, development, materials, decorative schemes, figured subjects and relationship to room functions. The discussion of each topic is necessarily brief, and may not satisfy everyone: for example, it might have been nice to have a fuller discussion of the relationship between mosaics and the spaces that they decorated, perhaps engaging with work such as that of Sarah Scott and Shelley Hales, which draws on recent developments in the study of domestic space and decoration in the wider Empire in order to understand Romano-British mosaics as products of their social context.Footnote ³ Nevertheless, there are many useful insights in these sections. Here (and elsewhere in the catalogue, e.g. at Brading, nos III, 331.1 and 331.6) the authors' encyclopaedic knowledge of the mosaicists' repertoire leads them to be sceptical of elaborate allegorical or cultic interpretations of what is often fairly commonplace imagery. Their intimate familiarity with the style and workmanship of the mosaics, derived especially from drawing them, also enables them to make an authoritative contribution to the debate over the nature and significance of the regional groupings originally identified by D.J. Smith,Footnote ⁴ and subsequently refined by Smith and others. In place of Smith's ‘schools’ or ‘workshops’, which imply more-or-less stable groups of craftsmen operating from a base in a major town, Cosh and Neal favour a more flexible model to account for the complex web of similarities and differences between mosaics, envisaging fluid groups of craftsmen who moved around a region, and occasionally beyond it, sometimes working separately, sometimes collaborating on large projects, and picking up ideas from each other in the process. This seems more likely than the traditional idea of ‘schools’, but it is still very much a craftsman-centred view, although they do allow a role for the patron, and his desire to emulate his neighbours, in producing the distinctive regional patterns (e.g. Vol. IV, 16). The analysis of stylistic groupings also yields insights into the economy of Roman Britain: for example, it suggests that in the fourth century mosaicists were brought into South-East England from other parts of the country to carry out commissions, presumably because there was too little demand to support local specialists (Vol. III, 21; IV, 27).

Some might question the value of such a comprehensive catalogue: do we really need to know about every fragment, no matter how small, and every mosaic noted in the seventeenth or eighteenth century and now lost without trace? This kind of empirical, artefact-centred research is somewhat unfashionable at present, and Neal and Cosh are open about their desire to continue the tradition of antiquarians such as Samuel Lysons.Footnote ⁵ But these volumes will serve current and future research agendas as well as more traditional ones. Including every mosaic, even the most crude or fragmentary, gives us a more representative picture of mosaic production than that suggested by the well-known but untypical figured pavements, and the corpus will thus help to broaden the focus of mosaic studies beyond traditional art-historical approaches, to address more social and economic questions. The catalogue provides a large enough sample to serve as a basis for quantitative analysis, which will be facilitated by the systematic provision of dimensions, tessera sizes, and so on. Some of the work outlined in the introductory and concluding sections of the present volumes starts to explore the possibilities offered by this rich collection of material: for example, an appendix in Volume III (560–5) compares the areas of rooms with mosaics and of their decorated panels in Colchester, Silchester, Verulamium and London, revealing a significantly different pattern in London, with a smaller average room size and a larger proportion of the floor left undecorated; the difference might be related to the commercial role of London, or perhaps to higher land values. But this is only scratching the surface; there is far more that can now be done. The inclusion of information about the architectural setting of the mosaics and of so many building plans means that the catalogue will also facilitate study of the social aspects of the mosaics and of Romano-British domestic life — quite apart from the mosaics, this must be one of the largest available collections of Romano-British house and villa plans. In short, these volumes should provide a springboard for research into a wide range of topics, not only in Britain, but also beyond: hopefully, by making such a huge amount of material easily available, they will help to remedy the marginalisation of Roman Britain in research on Classical art.

The other aspect of the work that might be questioned is the decision to illustrate the mosaics primarily through paintings rather than photographs. The paintings inevitably involve a degree of interpretation, especially where they are based on oblique black-and-white photographs combined with excavators' descriptions of the colours,Footnote ⁶ and the uniformity of style perhaps creates an unrealistic impression of uniformity in the mosaics. However, at least the interpretation is largely that of only two individuals, and unlike some of their predecessors, Neal and Cosh are careful not to conjecture beyond what actually survives; missing areas of the design are indicated in outline rather than restored. The paintings are undoubtedly clearer than photographs, where oblique angles, uneven focus, dirt and shadows often obscure the design, and raking light, reflecting unevenly off the tesserae, can make black areas appear white and vice versa. Their uniformity helps to bring out patterns in design and materials, such as the use of cream or grey limestone for plain areas in western and south-western England, in place of the red tile used elsewhere. The bright colours probably bring us closer to the original appearance of the mosaics, although the vertical perspective bears little relation to the way in which the mosaics would have been viewed and experienced; it might be interesting to see some of the paintings incorporated into digital reconstructions of the buildings.

Volume IV ends with a collection of brief biographical sketches of the individuals whose illustrations of mosaics appear in the corpus, from William Camden in the sixteenth century to Luigi Thompson in the twenty-first. Cosh and Neal modestly omit themselves, but they are worthy successors to Samuel Lysons and the other artists whose work they emulate. However, this is much more than an antiquarian exercise, and it sets a standard that the authors of other catalogues must aspire to, but will probably never match. It will form a solid foundation for research into Roman mosaics and the economy and society of Britain and the Empire far into the future.